Pages

Tuesday, November 29, 2011

Skimming : From Underdogs to Tigers

From Underdogs to Tigers: The Rise and Growth of the Software Industry in Brazil, China, India, Ireland, and Israel



The Indian Software Industries

  • The phenomenal success and growth of the Indian software industry is quite extraordinary when one considers that India is a very poor country with poor infrastructural investment (generally and in IT) and an illiteracy rate over 33 percent. (p.13)

Early Entry - pre-1984

  • The single most dramatic event of this period, however, was the departure of IBM in protest against the FERA rules, which required it to dilute its equity holding to 40 percent in 1977. The departure created an import substituting opportunity for domestic manufacturers of computers, and a demand for programmers that could write the proprietary software to run computers produced by domestic hardware firms. (p.22)
  • The exit of IBM also provided an opportunity for sales of (micro) computers by other foreign hardware manufacturers. Foreign companies such as Burroughs depended upon software programmers in India to write software conversion programs which could be used by non-Burroughs (mostly IBM) clients in order to switch to Burroughs computer systems. (p.22)
  • The big achievement of Indian firms operating in this period was that they developed the ability to compose a team of talented software programmers to deliver highly bespoke (technical) services on a variety of software systems to large foreign firms. Their relative specialization was porting data from one platform to another and maintaining legacy systems on old mainframe platforms. (p.23)

New Entry and Experimentation: 1985–91

  • Thus, the shift to networked computing in the West opened up a huge new source of demand for software services. These developments created a huge demand for software services in addition to the maintenance of legacy mainframe systems, the speciality of several Indian firms. There was demand for customized software that would allow firms to migrate from mainframe to networked systems and install and maintain enterprise resource systems. (p.24)
  • Indian firms, by now, had considerable experience working with the migration of data across different systems. Also, the increased credibility of Indian software professionals in the Western world created a positive externality that helped new entrants during this period. More importantly, it was in this period that foreign firms first realized the large cost advantage of employing Indian programmers. (p.24)

Imitative Entry and Financial Liberalization: 1992–99

  • The liberalization of financial flows from 1992 meant that foreign capital could move in relatively easily. This was particularly opportune for the Indian software industry. … The experience of COSL and TI had demonstrated that an Indian subsidiary of an MNC could operate as a low-cost outsourcing center for global software needs. … Therefore, software was first developed at the Indian subsidiary and then installed on-site by teams of Indian software professionals. (pl.26-27)
  • The dramatically decreasing cost of telecommunications access and its vast outreach due to the STPI scheme also meant that offshore operations came within the reach of smaller firms as well. Desai argues that offshore operations became popular as many firms switched to web-based delivery made possible by the STPI scheme. (p.27)
  • By the end of this period of unfettered demand, the Indian software industry had built both a general capability (for outsourced service delivery) and some firm-specific capabilities (in particular domains and software process management). (p.29)

Consolidation and Slowdown: 2000 Onwards

  • The slowdown in demand came on the heels of the dot com crash and the recession in the United States, the industry’s largest market. (p.30)
  • The outsourced business model required a certain organizational capability in human resource management and in software process management to ensure reliability of the service product.(p.33)


  • Firms responded in some measure by adopting procedural norms that would make their software development immune to such attrition. One of these was to rely heavily on documentation and another was to start using proprietary tools wherever possible. New organizational structures evolved in order to keep a mix of new and old employees as an informal deputizing system. (p.35)
  • The growing scale and offshore component of projects facilitated a careful splitting of tasks. Thus, in many offshore projects, process control allowed something like a ‘Babbage effect’ in the growing specialization of the industry in outsourced services. The best quality programmer could be used for suitable tasks while less able/experienced programmers could be assigned the lower-level tasks, each paid according to their productivity (p.35)
  • However, it is the winning combination of particular organizational capabilities and new business models that differentiate Indian software firms from other software providers in the global marketplace, allowing them to hold a fifth of the global market for custom software. (p.36)

The Irish Software Industry

  • Key to this turnaround, and also to the rise of the Irish software industry, was the implementation of a highly successful policy of foreign direct investment, which has resulted in Ireland becoming home to over 1,050 foreign companies, 46 percent of US origin. (p.42)
  • The origins of the IT industry in Ireland can be traced back to this time when the IDA made the decision to selectively target companies in potentially high-growth, high-tech sectors. This strategy resulted in a wave of foreign investment in the 1970s in electronics and computer hardware with Digital Equipment Corporation becoming the first company to establish a mini-computer manufacturing operation in Ireland in 1971 (p.43)
  • The success of the IDA in attracting both Microsoft and Lotus Development (now part of IBM) to Dublin in 1985 had a catalytic effect on the development of the Irish software industry. Both companies produce off-the-shelf products for mass markets and their Irish operations were initially responsible for the manufacturing and distribution of these products, including tasks such as duplicating disks, printing manuals, and assembling shrink- wrapped packages. (p.46)
  • Over time Ireland, unlike India, did not follow the trajectory of providing international software services, but rather developed a range of firms which sell software products in international niche markets in systems software and enterprise applications. (p.49)
  • Several Irish software companies started out by providing ‘bespoke’ or custom services to businesses, expanding this into consultancy kits and then products, which many then succeeded in exporting. (p.49)
  • Other Irish software companies came into existence either as a result of government initiatives or contracts (including spin-offs from state-owned entities) or when firms in other industries, such as telecommunications or computer hardware, spun off their software divisions. (p.50)
  • Several of Ireland’s most successful indigenous companies have their origins in the university environment. ...
    Ireland’s most successful indigenous software company, IONA Technologies can trace its roots to the EU Esprit grants for distributed computing research in the Computer Science Department of Trinity College, Dublin. (p.51)
  • The mid-1990s saw an upsurge in entrepreneurial start-ups many of whom are product oriented from the outset, focused on international markets, have obtained VC funding, and have the fortune to avail the managerial expertise of seasoned software industry veterans. (p.52)
  • Until 1990, indigenous firms were largely reliant on the provision of software services (mainly bespoke software development), but the mid-1990s, saw a major switch taking place from services to products, and from servicing the local market to exporting. (p.53)
  • As indigenous companies increasingly move toward product provision, they do so in niche (or vertical) markets. O’Malley and O’Gorman [6, p. 29] state that ‘indigenous producers of software products and services to a lesser extent, usually follow a strategy of specialization in market niches’. The reasons for this are twofold. First, this specialization implies firms can succeed without having to be very large, an important consideration given the limited size of the domestic market. Second, there can be relatively lower barriers associated with entering vertical markets, which often sees firms starting out by providing a custom-based product for a single customer and using this success to secure other customers in the same niche (often in Ireland or the United Kingdom), while at the same time developing their application into a generic product suitable for export. (p.57-58)
  • The Irish government did not set out with the express intention of developing an indigenous software industry, but they were highly successful in luring software MNCs and in generating the right input factors, notably overhauling the telecommunications system, creating a favorable business (tax) environment, and investing heavily in the national stock of human capital, particularly in science and technology disciplines. (p.67)

The Israeli Software Industry

  • Unlike the Indian software industry, the Israeli industry is product based and R&D intensive, with its four undisputed leaders—Amdocs, Comverse, Mercury Interactive, and Checkpoint Software Technologies—either inventing or leading their market niches. (p.72)
  • Indeed, software is just the latest manifestation of Israel’s comparative advantage in R&D. A comparative advantage that is evident when one compares Israel’s patents profile, one of the best proxies of industrial innovation, with other emerging countries. Not only Israeli organizations patent much more than their counterparts, but the quality of their patents and their scientific linkage is much higher (p.73)
  • Israel has always had strong academic involvement in industrial R&D (p.76)
  • The official history of IT and computing in Israel began before the creation of the Israeli state. In 1947, the advisory committee of the Applied Mathematics Department of the Weitzmann Institute (then known as the Seiff Institute), consisting of Albert Einstein, Hans Kramer, Robert Oppenheimer, John Von Neumann, and Abram Pais, recommended that the Institute build an electronic digital computer, making Israel the first state-to-be to commit itself to computing (p.79)
  • However, the rapid expansion of defense R&D and the rapid accumulation of IT skills by both university graduates and graduates of the military technological units created local demand for IT usage along with the knowledge base to supply it. (p.79)
  • All of these companies have been able to undertake large-scale R&D efforts due to R&D grants given by the Office of the Chief Scientist (OCS) of the Ministry of Trade and Industry (now Trade, Industry, and Employment) after the recognition of software as an industry in 1985. (p.80)
  • Israeli firms established during this period specialized in products that solved problems connected with IT usage. Therefore, it is not surprising that the original data security companies (for which Israel is world-renowned) were established at this time. (p.83)
  • As noted earlier, software is the latest manifestation of Israel’s comparative advantage in R&D. This comparative advantage lies in the high levels of skills and talent of Israeli scientists and engineers, and the relative paucity of other types of natural advantages, such as natural resources. (p.86)
  • But perhaps the most crucial decisions made in these years were the enactment of the R&D Law in 1984 and the recognition of software as an industrial sector in 1985. One of the main provisions of the R&D Law was that the OCS would not have a limited annual budget for its R&D grant scheme. Therefore, all legitimate and feasible proposals for R&D product-oriented projects would be sponsored. (p.89)

The Brazilian Software Industry

  • Lacking familiarity with English, a key export language, these large countries rely upon a strong local demand for the production of software through their economies’ commitment to Information and Communication Technologies (ICT). (p.103)
  • As suggested by Table 5.3 the leading activity in the software industry in Brazil is system integration, followed by processing services, and hardware and software support. (p.106)
  • Another important characteristic of the industry is the prevalence of government software firms among the major players. These government firms originated before the 1980s. (p.109)
  • In the decades preceding the 1990s, IT users, private and public sector alike, perceived software development as an essentially auxiliary activity performed in-house by IT-user organizations and hardware producers. That is, software production activity did not have either a corporate or an industrial identity and was mostly an afterthought, a marginal activity within firms. (p.111)
  • The Brazilian Payment System (SPB) is a particularly good example of how idiosyncratic local needs can spur the development of capabilities. In 1999, the Brazilian Central Bank decided that, by 2002, Brazil should have installed an advanced payment system in accordance with the most up-to-date Internet technology recommended by the Bank for International Settlements. (p.115)
  • Start-ups are a third area of interest in telecom in particular for cell phones. Demand for cell phones has grown at a double-digit rate over the past few years, reaching thirty-three million active phones by the end of 2002 and is expected to continue to grow in the future. (p.117)
  • Another area where the Brazilian government is a world leader is in the adoption of ‘free’ software such as Linux. At least eleven Brazilian cities have passed laws giving preference to or requiring the use of ‘software libre’ and a number of other municipalities, states, and the national government have considered similar legislation. (p.117)
  • The Brazilian government is very large, with sophisticated needs and requirements in terms of information processing. To meet these needs, the strategy has been to create public firms that supply software and IT services across states and sometimes across the entire country. These companies have a combined employment of over 10,000 and have developed a rich pool of competencies. This effort has made Brazil one of the leading countries in this area, with important flagship projects such as electronic voting and electronic tax declarations. (p.117)
  • Through the mid-1990s, Microsiga and Datasul, two of the largest Brazilian software firms, had the market to themselves. But liberalization attracted giants such as SAP and BAAN to the business opportunity the Brazilian market presented. While these global producers have quickly taken over the large software firm market (mostly from internal development, as this segment had never been the major target of local developers), the battle now is over ERP systems tailored to cater to medium-sized companies. (p.118)
  • Brazil is attracting the interest of leading Indian services firms such as Tata Consulting Services, which announced a joint venture with TBA, a local firm, to create a software development operation expected to grow to 3,000 professionals in five years. In early 2004, it was the first software firm in Brazil to receive CMM level 5 certification. (p.121)
  • Industry prospects changed when at least two conditions were established. One of these is the presence of lead domestic client sectors for software firms with demands close to those of leading international firms. These clients provide opportunities for learning and competence deepening similar to those found when exporting to foreign competitive firms. The examples of banking and telecom explored in Section 5 illustrate this dynamism. But the opportunity to learn is not always enough. The second required condition is the presence of competition and selection mechanisms, which induce successful firms to structure capabilities, while winnowing out firms that fail to learn. (p.122)
  • This focus of the software sector on the domestic market may also have important multiplying effects for Brazil. Recent studies at the company and country levels have shown that investment in information technology is positively related to corporate and national economic performance [48–50]. This is especially important for sectors that Kraemer and Dedrick [48] refer to as ‘production close to use’ sectors, where IT use is very close to or overlaps with production.

The Chineese Software Industry

  • China’s software firms can be loosely classified into two groups. The first, which includes many of China’s largest software firms, are systems integration (SI) firms. Systems integration involves large custom projects where the software firms provide a combination of hardware and software and services, including customized software. ...
    The second type of firm focuses primarily on product development, although for various reasons, also ended up doing SI and services. (p.133)
  • As with the Brazilian software industry, the Chinese software industry relies heavily on the huge market of users and domestic producers. The rapid and fairly diverse industrial growth of China has led to a large surge in domestic demand. (p.139)
  • In addition to having a direct influence on software R&D and the formation of software firms, the government was also deeply involved in how the hardware industry grew. This has ramifications for software firms in two ways. First, some of the larger hardware firms have developed software arms, or are combining software and hardware in systems integration activities. Second, some hardware manufacturers provide business to software firms, with needs for software (to run on their products) ranging from desktop applications to embedded software. (p.141)
  • In 1999, CASS designed its own Hopen embedded operating system, and has leveraged this to great effect in the electronics manufacturing sector.(p.142)
  • While government procurement in some product markets may have been more beneficial by way of getting some firms started, too much support may make them wholly dependent. Systems integrators that have relied too much on large government contracts may not be able to foster other customers, and so their growth may be limited in the future. (p.146)
  • Recently, the Hopen system was further developed as the NUWA project,21 in anticipation of the increasing market for embedded software. The NUWA project focuses on developing the NUWA-Hopen OS as a basic embedded software platform to work across a variety of information appliances. (p.151)
  • Since software is critical to their success, Huawei is making strategic investments in software, including a large development center with over 400 engineers in Bangalore. They set up shop in Bangalore as opposed to developing their software in China suggesting that in the medium term at least, Chinese software process capability, even in embedded applications, may not be as good as that of India’s. Our interviews with Huawei executives in Bangalore showed that they believed that India’s process capability was superior to China’s, even though the Chinese were more advanced in terms of telecoms domain knowledge and had at least as much knowledge of software theory. (p.151)
  • In 2000, the largest user of ERPs was the manufacturing industry, at 68.2 percent of the total ERP market, followed by the distribution, transportation, telecom, and health industries. (p.152)
  • Many enterprises value service more than anything else, and require ERP vendors to have strong customer support, something Chinese vendors are more prepared to do than multinationals, which would rather just sell a product with minimal additional cost of localization. (p.153)
  • Chinese software firms compete heavily with each other at the ‘lower’ end of areas such as office automation and basic ERPs, but revenues are harder to build up because these clients are less sophisticated or well-off and tend to require uniquely developed projects or much handholding (i.e. consulting or customer service) which works against the software firms achieving scale economies. (p.153)
  • These numbers provide a powerful reason why so many software companies— product firms and systems integrators alike—are doing systems integration. In our interviews, many firms, including product firms, noted that it was difficult to survive on products alone. Even Digital China, the software arm of Legend, focused on SI in order to gain additional revenue (as related to us by one of its executives). (p.154)
  • One problem is that even though local software firms have a lot of work requiring highly customized, domestic oriented systems and systems integration work, it is difficult to derive scale economies in such uneven markets. A second problem that Chinese software firms face is the problem of high competition. Many local firms compete with each other at the low end. At the higher end of the market in certain sectors, there is also the increasing presence of foreign firms. (p.155)
  • However, while Chinese software workers have the theoretical knowledge to handle the basic tasks and are fairly sophisticated analytically, they lack the software process skills of Indian software workers. (p.156)
  • Various younger companies such as Tangram, which makes educational system software, and Intrinsic Technology, which makes mobile device software, noted that it was difficult to develop and grow because they lacked managers who could implement the company’s vision. (p.157)
  • Piracy is also often mentioned as something that continues to hurt the software industry, and is now being officially recognized as a problem. Firms that we interviewed recognized that many software pirates are simply firms or individuals who cannot afford the software, and whom the state may be unwilling to penalize. (p.158)
  • At least three distinct strategies appeared to be adopted in order to address the need to survive and grow under these conditions. The first strategy is that of specialization. Specialization is increasingly seen in the better product firms such as Kingdee, Kingstar, and Red Flag. Specialization is one way of deepening competencies and differentiating oneself from other firms that cannot do so. (p.159)